Methods and processes for non-invasive assessment of genetic variations

Abstract

Provided herein are methods, processes, systems and machines for non-invasive assessment of genetic variations. In particular, provided herein are methods, processes, systems and machines for non-invasive assessment of copy number variations. In some aspects, copy number variations include aneuploidies (e.g., trisomy 13, 18, or 21). In some aspects, copy number variations include microdeletions or microduplications.

What is claimed is: 
     
         1 . A method for reducing bias in sequence reads for a test sample comprising:
 (a) generating a relationship between (i) local genome bias estimates for sequence reads of a test sample and (ii) bias frequencies for the sequence reads of the test sample, thereby generating a sample bias relationship, wherein:   the sequence reads are of circulating cell-free nucleic acid from the test sample, and   the sequence reads are mapped to a reference genome;   (b) comparing the sample bias relationship and a reference bias relationship, thereby generating a comparison, wherein,   the reference bias relationship is between (i) local genome bias estimates for a reference and (ii) bias frequencies for the reference; and   (c) normalizing counts of the sequence reads for the test sample according to the comparison generated in (b), whereby bias in the sequence reads for the test sample is reduced.   
     
     
         2 . The method of  claim 1 , wherein each of the local genome bias estimates is determined by a process comprising use of a kernel density estimation. 
     
     
         3 . The method of  claim 1 , wherein each of the local genome bias estimates for the reference bias relationship and the sample bias relationship is a representation of local bias content. 
     
     
         4 . The method of  claim 3 , wherein the local bias content is for a polynucleotide segment of 5000 bp or less. 
     
     
         5 . The method of  claim 1 , wherein each of the local genome bias estimates is determined by a process comprising use of a sliding window analysis, wherein:
 the window is about 5 contiguous nucleotides to about 5000 contiguous nucleotides and the window is slid about 1 base to about 10 bases at a time in the sliding window analysis; or   the window is about 200 contiguous nucleotides and the window is slid about 1 base at a time in the sliding window analysis.   
     
     
         6 . The method of  claim 1 , wherein the comparing in (b) comprises generating a fitted relationship between (i) ratios, each ratio comprising a bias frequency for the test sample and a bias frequency for the reference, and (ii) local genome bias estimates. 
     
     
         7 . The method of  claim 6 , wherein the fitted relationship in (a) is obtained from a weighted fitting. 
     
     
         8 . The method of  claim 1 , wherein the normalizing in (c) comprises factoring one or more features other than bias, and normalizing counts of the sequence reads. 
     
     
         9 . The method of  claim 8 , wherein the factoring one or more features is by a process comprising use of a multivariate model. 
     
     
         10 . The method of  claim 8 , wherein counts of the sequence reads are normalized according to the normalizing in (c) and the factoring of the one or more features. 
     
     
         11 . The method of  claim 1 , comprising, after (c), generating a read density for one or more portions of a genome, or a segment thereof, according to a process comprising generating a probability density estimation for each of the one or more portions comprising the counts of the sequence reads normalized in (c). 
     
     
         12 . The method of  claim 11 , wherein the probability density estimation is a kernel density estimation. 
     
     
         13 . The method of  claim 11 , comprising generating a read density profile for the genome or the segment thereof, wherein the read density profile comprises the read densities for the one or more portions of the genome, or the segment thereof. 
     
     
         14 . The method of  claim 11 , further comprising adjusting each of the read densities for the one or more portions. 
     
     
         15 . The method of  claim 14 , wherein the adjusting is according to a weighting factor wherein adjusting each of the read densities for the one or more portions according to a weighting factor comprises adding, subtracting, multiplying and/or dividing a read density by a weighting factor. 
     
     
         16 . The method of  claim 15 , wherein:
 the weighting factor is determined according to a regression, and   the regression results from a comparison between bias frequencies of local genome bias estimates of the reference and the test sample.   
     
     
         17 . The method of  claim 1 , wherein:
 the local genome bias estimates for the sequence reads of the test sample comprise local GC densities, and   the bias frequencies for the sequence reads of the test sample comprise GC density frequencies.   
     
     
         18 . The method of  claim 1 , wherein:
 the local genome bias estimates for the reference comprise local GC densities, and   the bias frequencies for the reference comprise GC density frequencies.   
     
     
         19 . The method of  claim 1 , wherein (a), (b), and (c) are performed by a microprocessor. 
     
     
         20 . The method of  claim 1 , wherein thousands to millions of sequence reads are mapped to the reference genome.